Bone Changes with Long-Acting Cabotegravir or Tenofovir Disoproxil Fumarate/ Emtricitabine for HIV Prevention in Cisgender Men and Transgender Women: HPTN 083

Abstract

In a randomized clinical trial, pre-exposure prophylaxis (PrEP) with long-acting cabotegravir (CAB-LA) had a better bone safety profile than tenofovir disoproxil fumarate-emtricitabine (TDF-FTC) over 105 weeks. For individuals with low bone mineral density or other fracture risk factors, CAB-LA PrEP should be considered over TDF-based PrEP.

Oral pre-exposure prophylaxis (PrEP) with tenofovir disoproxil fumarate-emtricitabine (TDF-FTC) is highly effective for HIV prevention, but its efficacy is limited by incomplete adherence¹. TDF-FTC also has been shown to decrease bone mineral density (BMD)^2,3, and its effects may not be fully reversible upon discontinuation⁴, raising safety concerns for long-term usage. In HPTN 083, long-acting cabotegravir (CAB-LA) was found to be superior to TDF-FTC for HIV prevention among cisgender men and transgender women who have sex with men (MSM/TGW)⁵. However, the relative bone safety of these regimens is unknown.

METHODS:

HPTN 083 is an ongoing randomized, double-blind, double-dummy, noninferiority trial comparing CAB-LA (600 mg, given intramuscularly every 8 weeks) to daily oral TDF-FTC (300 mg-200 mg) for the prevention of HIV infection in MSM/TGW⁵. Eligible participants were adults (≥18 years of age) who had undetectable blood HIV RNA viral load within 14 days before trial entry and negative HIV serologic tests, were in general good health, including normal kidney function, and were likely to be exposed to HIV. Key exclusion criteria were the use of illicit intravenous drugs within 90 days before enrollment, previous participation in the active treatment group of an HIV vaccine trial, coagulopathy, buttock implants or fillers, seizure disorder, or a corrected QT interval of greater than 500 msec. Participants who had positive results on a hepatitis B virus surface antigen test or hepatitis C virus antibody test were also excluded. The Dual-energy X-ray Absorptiometry (DXA) substudy was offered to participants at 19 sites where Hologic or Lunar DXA instruments were available for a target enrollment of 350 participants. Participants who were > 300 lbs were not eligible for the DXA substudy. The study (clinicalTrials.gov identifier NCT02720094) was approved by the institutional review board, ethics committee, ministry of health, or a combination of these entities at each participating site. All participants provided written informed consent.

BMD was assessed by DXA at the lumbar spine (L1-L4; LS), femoral neck (FN), and total hip (TH) at baseline, 57 weeks, and 105 weeks. Sites were instructed to use the same scanner and the same hip (left) at both study time points for each substudy participant. All scans were read centrally by staff blinded to treatment assignment and clinical characteristics, using a standardized protocol at the Body Composition Analysis Center, Tufts University (Boston, MA, USA).

The primary objective of this analysis was to compare the percentage change in BMD at the three anatomical sites from baseline to week 57 and baseline to week 105 in those randomized to CAB-LA versus TDF-FTC PrEP. We also examined the proportion who experienced clinically significant BMD loss (> 5%) over the study interval. Participants were included in this analysis if they received at least ten bi-monthly injections over the first 18 months from enrollment (i.e, as-treated population). A sensitivity analysis was conducted in all substudy participants, regardless of PrEP dose. Two-sample, independent t-tests (assuming unequal variances, Satterthwaite method) were used to compare the mean difference between each study arm. Differences between the arms in the proportions of those with BMD > 5% bone loss was determined using Chi-square tests.. All statistical analyses were performed with SAS (Version 9.4, Cary, NC).

RESULTS:

Of the 1372 participants enrolled from HPTN 083 sites participating in the DXA substudy, 351 (26%) (314 MSM and 37 TGW) were enrolled in the DXA substudy. Of these, 22 did not receive any injections during the study period and 75 (41 in TDF-FTC arm and 34 in CAB-LA arm) received fewer than 10 bi-monthly injections. Of the 254 in the as-treated population (230 MSM and 24 TGW), 47% were white, 31% were Black, and 12% were Asian, balanced across randomized arms. At baseline, the median (Q1, Q3) age was higher in the TDF-FTC arm compared to the CAB-LA arm (29 (23, 41) v 26 (22, 31) years; p=0.03). Median (Q1, Q3) body mass index was similar between the arms (26.5 (22.8, 29.6) vs 25.1 (21.9, 29.2) kg/m², p=0.29). (Supplemental Table 1).

At 57 weeks (n=119 TDF-FTC; n=129 CAB-LA), the mean (standard deviation (SD)) percentage change from baseline in LS BMD was −0.93% (3.6%) in the TDF-FTC arm and 0.89% (3.1%) CAB-LA arm (between group p<0.01). The mean (SD) percentage change in femoral neck BMD was −1.13% (4.9%) in the TDF-FTC arm and 0.57% (3.9%) CAB-LA arm (between group p<0.01). The mean (SD) percentage change in total hip BMD was −1.13% (4.2%) in the TDF-FTC arm and 0.46% (2.9%) in the CAB-LA arm (between group p<0.01). At 105 weeks (n=96 TDF-FTC; n=107 CAB-LA), the mean (SD) percentage change from baseline in LS BMD was −0.78% (4.9%) in the TDF-FTC arm and 1.65% (4.3%) in the CAB-LA arm (between group p<0.01). The mean (SD) percentage change in femoral neck BMD was −1.0% (6.3%) in the TDF-FTC arm and 0.50% (4.9%) in the CAB-LA arm (between group p=0.07). The mean (SD) percentage change in total hip BMD was −0.59% (5.4%) in the TDF-FTC arm and 0.79% (3.3%) in the CAB-LA arm (between group p=0.03) (Figure). The magnitude of BMD changes in each of the treatment arms was similar in participants < 25 years compared to ≥ 25 years (Supplemental Table 2). At 105 weeks, the proportion of participants with >5% loss of BMD from baseline was greater in the TDF-FTC arm compared to the CAB-LA arm (LS: 13.9% vs 3.8%, p<0.01; FN: 15.6% vs 9.9%, p=0.17; TH: 9.8% vs 3.8%, p=0.05; >5% loss at any of the 3 sites: 24.6% vs 15.2%, p=0.06). Results of analyses conducted in all substudy participants, regardless of PrEP dose were similar (data not shown).

Figure.

Percentage Bone Mineral Density Change over 105 weeks in Individuals Initiating PrEP with either Long Acting Cabotegravir or Tenofovir Disoproxil Fumarate (TDF)/Emtricitabine (FTC)

DISCUSSION

Currently approved PrEP formulations by the US Food and Drug Administration include oral preparations (TDF-FTC, and tenofovir alafenamide (TAF)-FTC, for non-vaginal sexual exposures only) as well as long-acting cabotegravir. The selection of a PrEP regimen is highly individualized based on patient preference, cost, availability, mode of administration, and potential adverse effects. TDF exposure in the setting of PrEP has been associated with about a 1% net decrease in BMD over 48 weeks^2,3.

Similar to these findings, our study showed a BMD loss in the TDF-FTC group of 0.93–1.13% at 57 weeks and 0.59–1.0% at 105 weeks. This BMD loss is of unclear clinical significance in that it is relatively small in magnitude and generally reversible with TDF discontinuation⁶. However, a recent systematic analysis of seven randomized clinical trials showed a trend towards increase risk of fracture in persons on TDF-based PrEP with a relative risk 1.24; 95% confidence interval 0.97, 1.56)⁷. Longer-term longitudinal studies are needed to better understand whether TDF-based PrEP is associated with fractures, and which populations are most at risk. Among adolescent cisgender males who have yet to reach peak bone mass, discontinuation of TDF-based PrEP may not lead to full recovery of age-appropriate BMD⁴, potentially increasing the risk of fracture later in life. In addition, the risk of fracture related to TDF-based PrEP in older individuals (eg > 50 years), those with concomitant medical conditions/treatments associated with fracture, or those with other risk factors for fracture (eg family history of fracture), should be carefully examined. The efficacy of concomitant vitamin D supplementation to attenuate bone loss in the setting of TDF-based PrEP, which was suggested by a prospective study examining bone turnover markers⁸, also requires further investigation.

Compared to the TDF-FTC arm, the MSM/TGW randomized to CAB-LA had a significant increase in BMD over the study interval. An increase of similar magnitude in BMD over the study interval was also seen the TAF-FTC arm of the DISCOVER trial which compared TDF-FTC to TAF-TFC for HIV prevention in a similar population⁹. We explored whether this effect was driven by participants < 25 years, who are more likely to have not reached peak bone mass; the median percentage change in BMD were similar in these age groups. It is unclear whether this increase in BMD is related to changes in health-related behaviors (eg, increased physical activity) accompanying participation in a clinical trial, i.e. a Hawthorne effect. Nonetheless, our findings suggest that CAB-LA and other non-TDF regimens, such as TAF-FTC, are important options for MSM/TGW wishing to avoid potential bone toxicity of TDF-based PrEP. It is not known whether our results extend to cis-gender women. According to CDC guidelines¹⁰, individuals at higher risk of fractures receiving or potentially initiating TDF-based PrEP should be referred for “appropriate consultation and management.” In this population, given its high efficacy for HIV prevention and bone safety, CAB-LA should be considered.